To Softmax, or not to Softmax: that is the question when applying Active Learning for Transformer Models

Open in new window